Preventing Overdose Using Information and Data from the Environment (PROVIDENT): protocol for a randomized,...

Area of Science:

• Public Health
• Data Science
• Epidemiology

Background:

• The accelerating drug overdose epidemic necessitates innovative strategies for resource allocation.
• Identifying high-risk communities is crucial for prioritizing public health interventions.

Purpose of the Study:

• To develop and evaluate PROVIDENT (Preventing Overdose using Information and Data from the Environment), a machine learning tool.
• To predict future overdose deaths at the census block group level.
• To guide geographically targeted public health resource allocation.

Main Methods:

• A randomized, population-based, community intervention trial in Rhode Island.
• Development of an interactive, web-based tool visualizing machine learning-based predictions.
• Randomized assignment of 39 municipalities to intervention (PROVIDENT) or control arms.

Main Results:

• Intervention efficacy assessed by comparing fatal and non-fatal overdose rates between treatment and control groups.
• Utilized Poisson or negative binomial regression for incidence rate ratio estimation.
• Municipalities in the treatment arm received neighborhood risk predictions to direct resources.

Conclusions:

• Predictive modeling can enhance public health decision-making.
• Geographically targeted resource allocation can improve overdose prevention and response.
• Findings will inform strategies for prioritizing communities for essential services.